Dimensions of Regional Crime in Türkiye and Their Socioeconomic Determinants: A Factor-Based Panel Data Approach
VII. International Applied Statistics Congress (UYIK – 2026), İstanbul, Türkiye, 11 - 13 Mayıs 2026, ss.135, (Özet Bildiri)
- Yayın Türü: Bildiri / Özet Bildiri
- Basıldığı Şehir: İstanbul
- Basıldığı Ülke: Türkiye
- Sayfa Sayıları: ss.135
- Bilecik Şeyh Edebali Üniversitesi Adresli: Evet
Özet
Crime is shaped jointly by individual behavior and the broader socioeconomic environment in which it occurs. Despite this complexity, much of the empirical work on the socioeconomic roots of crime collapses heterogeneous offenses into a single composite measure. Such a treatment tends to mask meaningful differences between offense categories and may bias estimated coefficients. This study takes a different route: rather than relying on predefined legal labels or a unified crime index, it uncovers the latent structure underlying twenty-one offense types and examines how socioeconomic conditions relate to each underlying dimension separately. The analysis draws on annual data for the twenty-six secondlevel statistical regions of Türkiye, organized as a balanced panel. A factor-analytic procedure groups the offense series into three coherent dimensions, broadly corresponding to violence and offenses against social order, acquisitive offenses, and smuggling-related activity along border zones. Scores derived from this structure are then used as outcome variables in panel regressions that absorb regional and yearly heterogeneity simultaneously, with inference adjusted for cross-sectional dependence and serial correlation. A spatial autoregressive specification serves as a robustness check. The estimates reveal that socioeconomic forces do not press uniformly on every dimension. A rise in unemployment translates into more acquisitive crime but leaves the other two categories untouched, whereas higher regional income appears to dampen violent offenses without affecting the rest. Income inequality, however, behaves differently: its effect is positive and meaningful for every dimension considered. The spatial diagnostics tell a similarly differentiated story; cross-regional spillovers are detectable for violent offenses alone, while acquisitive and smuggling-related crimes remain anchored to local conditions. These patterns argue against treating regional crime as a single object of intervention and point instead toward tailored, dimension-specific responses that also reckon with geographic interdependence.